COMBINATION OF SHS AND MECHANOCHEMICAL

SYNTHESIS FOR NANOPOWDER TECHNOLOGIES

T. Grigorieva, M. Korchagin, N. Lyakhov

Institute of Solid State Chemistry, 630128 Novosibirsk, Russia

Introduction

The synthesis of nanocrystalline mixed oxides is among the major problems of advanced ceramic technology. Conventional processes for manufacturing multicomponent mixed-oxide ceramics involve high-temperature reactions between metal oxides or between metal oxides and carbonates. These diffusion-controlled processes require applying high temperatures and using highly disperse precursor powders; mixed oxides are formed in particulates 1 to 5 m in size. Mechanochemical alloying partly removes the diffusion control, allowing for control from the thermodynamic parameters of the system. For example, for most reactions of mechanochemically assisted metal oxidation in a gaseous phase, the oxidation rate correlates with variations in Gibbs free energy [1]. For mechanically alloyed metals with negative enthalpies of mixing, the formation rate of solid solutions is a function of the enthalpy of formation of intermetallic compounds in equilibrium [2].

Therefore, if reactions with small decreases in Gibbs free energy are used in thermodynamic-controlled mechanochemical synthesis, a large power supply and long mechanical activation times would be necessary, by analogy with mechanochemical synthesis of mixed oxides from binary oxides, either alone or mixed with carbonates [3]. Efficient mechanochemical synthesis is not feasible unless in energy-intensive activators (steel drums and steel balls), and contamination of the product is inescapable. Therefore, mechanochemistry is almost useless for commercial ceramic processes with their extremely high requirements with respect to purity. The mechanochemical approach becomes practicable only with high rates of promoted reactions, i.e., those that are not only thermodynamically allowed, but also give an energy gain. High rates of mechanochemical processes render the final products highly dispersed, which can significantly influence the properties of the resulting ceramics.

One of the alternatives to mechanochemistry may be self-propagating high-temperature synthesis (SHS), which is energy-saving and can be applied to a large-scale production of many intermetallic compounds or complex oxides (nitrides, carbides) 4. Nevertheless, this method includes combustion stage, which requires very high temperatures. As a result, final products can be obtained only in the form of dense sintered or solidified products when SHS goes through the melting of reagents and / or products. To transform such products into commercially interesting powders, one needs to use milling as an unavoidable step. The grinding leads to a contamination again and to additional energy consumption, which may be comparable with that required for mechanochemical synthesis itself.

Another disadvantage of SHS, in many cases, is a non-uniform phase composition of products obtained due to high temperature gradients in a combustion wave resulting in quenching of some non-equilibrium phases.

The shortcomings of the two methods can be overcome and their ranges can be broadened by uniting the possibilities of these two methods; this allows one to obtain the final product of required phase composition and disperse state thus enabling the transition to more up-to-date resource-saving waste-free technology.

Experimental

The materials used in the investigation were: powdered tungsten and molybdenum, barium peroxide, tungsten and molybdenum oxides.

A ball planetary mill AGO-2 [5] was used for investigations. The volume of mill drums was 250 cm3, the ball diameter was 5 mm, the ball load was 200 g and the weighed portion of powder treated was 10 g. In order to avoid oxidation of metals, all the experiments on mechanical alloying were carried out in argon; initiation was carried out using an ignition which was lit by electric snail made of Nichrome. X-ray phase analysis was performed with a DRON-3M diffractometer with CuK radiation.

IR absorption spectra were recorded with a Specord 75 IR spectrometer; samples were prepared as tablets with annealed potassium bromide.

Electron microscopic studies were carried out using the JSM-T20 electronic microscope, and high-resolution electronic microscopes JEM-2010 and JEM-400.

The change of Gibbs' energy in synthesis of complex oxides from simple ones shows that all these reactions are thermodynamically allowed, while the majority of reactions with the participation of carbonates are available only at high temperatures (Table 1).

Some of these reactions were performed mechanochemically when a sufficient amount of energy was applied [3]. All the values obtained for systems under consideration are approximately of the same order of magnitude, so the conditions for these reactions to proceed should be rather similar. They cannot give bases to decide whether mechanochemical synthesis is possible.

Table 1

Changes of Gibbs' energy in the reactions of the mixtures of metal oxides and the metal oxides with metal carbonates resulting in the formation of complex oxides

No. / Reaction product
MeO + MeCO3 / Go298 /  Go1273
kcal/mol
1 / BaTiO3 / 16.3 / -17.7
2 / BaZrO3 / 24.2 / -10.0
3 / BaHfO3 / 21.9 / -13.2
4 / BaMoO4 / 2.5 / -30.5
5 / BaWO4 / -4.2 / -42.6
6 / BaAl2O4 / 26.5 / -16.3
7 / CaMoO4 / -9.3 / -44.7
8 / SrMoO4 / -6.6 / -41.7
9 / PbMoO4 / -1.6 / -35.6

The use of peroxide compounds and oxides as the initial components of mixtures, as proposed by some authors [3], does not bring substantial changes to the G of the reaction of complex oxide synthesis.

Among the methods to obtain oxides, energetically most profitable ones are the direct oxidation of metals (or metal mixtures) by oxygen (Table 2). However, in reality one can hardly perform this synthesis mechanochemically due to the high plasticity of metals.

Table 2

Changes of Gibbs' energy in the oxidation of metals or metal mixtures by oxygen or barium peroxide resulting in the formation of oxides and complex oxides

N / Product of reaction
Me + O2 / Gо298
kcal/mol / Product of reaction
Me + Me + O2 / Gо298
kcal/mol / Product of Ме + BaO2 interaction / Go298
kcal/mol
1 / BaO / -251.1 / Ba2TiO4 / -231
2 / TiO2 / -212.1 / BaTiO3 / -373.9 / BaTiO3 / -235
3 / ZrO2 / -248.5 / BaZrO3 / -402.0 / BaZrO3 / -263
5 / MoO3 / -159.7 / BaMoO4 / -334.9 / BaMoO4 / -196
6 / WO3 / -182.6 / BaWO4 / -364.5 / BaWO4 / -225
7 / Al2O3 / -378.0 / BaAl2O4 / -556.3 / BaAl2O4 / -390

The most promising way seems to be the oxidation of metals by peroxide compounds, especially if we take into account that there is rather large class of stable metal peroxides that can provide a couple for a proper metal to synthesize a complex oxide. In such cases, one should keep in mind that the mechanochemical interaction of metal peroxide with metals can follow two routes: either with the formation of complex oxides or with the formation of a mixture of simple ones, because both reactions are profitable from the thermodynamic viewpoint, and the change of Gibbs' energy is much higher than that for the synthesis from oxides and carbonates. Calculation of Go298 for the BaO2 interaction with metals demonstrated that the decrease of Gibbs' energy was larger by 30-40 kcal/mol than that when the products of the same reaction would be a sum of simple oxides. One can assume that synthesis will proceed to the formation of complex oxides.

According to X-ray diffraction data, mechanochemical interaction of barium peroxide with titanium for 5 min results in the formation of a mixture of barium titanates. Reflections corresponding to BaO2 and Ti disappear. There appear reflections corresponding to Ba2TiO4, BaTiO3 [6], etc. No diffraction reflections from simple oxides of barium and titanium are observed.

For mechanochemical interaction between barium peroxide and zirconium, reflections related to both initial reagents disappear within 5 min, while reflections related to barium zirconate BaZrO3 appear [7] along with some unidentified peaks (d = 3.64, 3.18, 2.667, 2.607, 2.309, 2.023, 1.933, 1.642).

In the interaction between barium peroxide and aluminium during mechanical activation for 5 min, both BaO2 and Al disappear completely; complex oxide BaAl2O4 appears [8].

The X-ray diffraction data from the products of mechanochemical BaO2 + Me reactions showed a series of unassigned reflections in addition to reflections from mixed oxides. It was not ruled out that these additional reflections were due to an unidentified oxide phase. The IR spectral characteristics for binary oxides and mixed oxides are well established. The spectral ranges of M-0 absorptions have been established exactly for various types and modifications of oxides.

Therefore, we carried out an IR spectral study of the products of mechanochemical BaO2 + metal reactions. The IR spectrum of the stock does no display absorption bands in the range of wavenumbers above 400 cm-1, whereas the systems examined exhibit M-0 absorption between 400 and 1000 cm-1.

For the BaO2 - Ti mixture, an antisymmetric absorption band peaking at 700 cm-1 and a very weak peak at 775 cm-1 appear after 1 min of activation. After 5 min of activation, a shoulder of the 700 cm-1 band appears at 550 cm-1 (Fig. 1b). This band is assignable [9-11] to  (Ti-O) stretches in [Ti04] tetrahedra of barium titanates [11]. The positions of the peaks and the band shape do not correspond to  (Ti-O) vibrations in various TiO2 polymorphs [10, 12 ].

The BaO2 - Zr mixture mechanically activated for 5 min exhibits an absorption band with a peak at 550 cm-1 (Fig. 1c). The peak position and shape of this band differ from those observed in the spectra of zirconium oxide polymorphs. Since barium -oxide absorbs in the spectral range below 400 cm-1 [9, 11], the 550 cm~1 band can be assigned to barium zirconate.

/ A broad asymmetric band with maximum at 530 cm-1 appears as a result of activation of a mixture of BaO2 with aluminium(Fig.1d). Magnesium aluminate exhibits a similar IR spectrum in this region [9, 13, 14] and the formation of barium aluminate can be assumed in the BaO2 - Al system under mechanical activation [15]. Electron microscopic investigation of the product of mechanochemical synthesis in a mixture of BaO2 with Zr shows that activation for 5 min results in the formation of particles 0,31 m in size. They are composed of small blocks 612 nm in size (Fig. 2). Microdiffraction picture obtained from a separate particle also points to a developed microblock structure.
Fig. 1. IR spectra of (a) initial mixture of barium peroxide with metals and after 1 min mechanical activation: (b) - Ti, ( c) - Zr, (d) - Al

Diffraction spots are shaped as rings, which is characteristic of polydisperse materials,while microdiffraction from a separate block gives a point, which is an evidence of single crystal state of the substance. Diffraction reflections in the both electron diffraction patterns, though somewhat broadened, are point reflections, what is the evidence of rather high degree of crystallinity of the substance formed in mechanochemical synthesis. A similar picture was observed also for other complex oxides, there is only a small difference in microblock size.

Fig. 2. Microphotograph of BaZrO3 particle obtained by mechanochemical synthesis during 1 minute

Thus, investigations demonstrate that the mechanochemical interaction of barium peroxide with metals, which proceeds with a substantial decrease of Gibbs free energy, as it follows from thermodynamic calculations, allows one to synthesize complex oxides with nanocrystalline particles at rather short time.

Mechanochemical oxidation of metals by peroxide compounds allows to prepare single-phase particles of complex oxides with nanometer-sized microblocks [16-19]. However, in case of some metals, such as tungsten, molybdenum, tantalum, mechanochemical synthesis does not proceed till completeness regardless of whatever reagents ratio would be, even after mechanical activation for a long time. Part of metal always remains unreacted. At the reagents molar ratio BaO2 : W = 1 : 1, the products are BaWO4 and Ba2WO5; unreacted W remains. At the increased barium peroxide content (to 2 : 1), part of W also remains unreacted. A mixture of tungstates BaWO4, Ba2WO5 and Ba3WO6 is formed. At the reagents ratio of 3:1, only the content of complex oxide Ba3WO6 increases. Mechanochemical activation of the mixtures of BaO2 with Mo, molybdates BaMoO4, Ba2Mo5 and Ba3MoO6 are formed. Similarly to the case of tungsten, a part of molybdenum remains unreacted.

SHS process in these mixtures also results in the formation of a mixture of complex oxide phases; unreacted metal remains. Preliminary mechanical activation of mixture has practically no effect on phase composition of SHS products though it increases the process rate.

The results obtained are the evidence that neither mechanochemically nor by SHS can one obtain single-phase product by the oxidation of a refractory metals by barium peroxide.

The interaction of the peroxides of alkali-earth metals with tungsten and molybdenum oxides, whose melting points are much lower than those of the corresponding metals, is also accompanied by substantial heat evolution.

IR spectroscopic investigations of the mechanochemical interaction of BaO2 with WO2 showed that the formation of complex oxide starts within the first seconds of activation. The IR spectra of the initial mixture BaO2 + WO2 contain one broad band at 850550 cm-1,without clearly exhibited maximums. It is assigned to the stretching vibrations  (W-O) (Fig. 3а) [9]. The  (Ba-O) band is below 400 сm-1.

/ After activation for 10 s, a clear band with maximum at 810 cm-1 is observed instead of the above-mentioned broad band; the intensity of that new band increases with increasing activation time (Fig. 3b, c). Similarity of the IR spectra of activated mixtures to the spectra of stolzite [9] allows us to assume that the activation of BaO2 + WO2 mixture results in the formation of BaWO4 with spinnel structure. However, according to the IR spectroscopic data, mechanochemical reaction BaO2 + WO2 BaWO4 does not proceed till its completeness, which is evidenced by the presence of noticeable absorption in the region 800500 cm-1 as a shoulder of the band with maximum at 810 cm-1 related to  (W-O) of the lower tungsten oxide.
Fig. 3. IR spectra of initial mixture of barium peroxide with WO2 (a), after its mechanical activation for 10 s (b) and 2 min (c), of SHS product in this mixture (d)

According to XRD data, a growth of the reflections from BaWO4 phase starts at the second minute of activation and reaches maximum by 5 minutes (Fig. 4 a, b). But then, with increasing time of activation, the intensity of diffraction peaks of the complex oxide does not increase and the intensities of peaks related to the initial tungsten oxide do not decrease. This means that the result of mechanochemical interaction BaO2 + WO2 is a mixture of phases.

/ For the interaction of BaO2 with MoO2, IR spectra and XRD also reveal a mixture of phases.
Rather large heats of formation of the complex oxides in the system involving the oxidation of the lowest tungsten oxide by barium peroxide allow one to perform these reactions by means of SHS. However, pretreatment of barium peroxide and sometimes heating of the initial mixture are necessary [20]. This is due to the fact that BaO2 particles kept in air get coated with a layer of barium carbonate and hydroxide. This layer prevents SHS reaction. One can assume that a short-term preliminary mechanochemical activation of the initial mixture leads to the destruction of these barrier layers and provides a substantial increase of the area of contact between oxides and barium peroxide.
Fig. 4. X-ray patterns of initial mixture of barium peroxide with WO2 (a), after its mechanical activation for 5 s (b) and of SHS product in this mixture (c)

Investigation of the effect of preliminary mechanochemical activation of commercially available BaO2 with WО2 showed that maximal temperature and rate of SHS process are achieved after activation for 2 min, and the single-phase complex oxide BaWO4 is formed (Fig. 3d and 4c). A disadvantage of the resulting substance was large particle size (300500 nm) and in some places even partial agglomeration (Fig. 5a). Electron microscopic studies show that subsequent mechanical treatment of the product for 2 minutes in high-energy activator of planetary type gives a material with particle size 2030 nm (Fig. 5b) without changes of phase composition (!).

Fig. 5. Microphotograph of SHS product of BaO2 + WO2 mixture (a), the same product after mechanical activation for 2 min (b)

Similar results were obtained also for the system BaO2 + MoO2, in which maximal rate of the SHS process with the formation of the complex oxide BaMoO4 is achieved after preliminary mechanical activation for 30 s.

Structural similarity of the higher oxide WO3 and barium tungstate BaWO4 and rather large heat of the reaction BaO2 + WO3 BaWO4 + 1/2 O2 (128 kJ/mol) allow us to assume that this reaction can be conducted mechanochemically. It follows from the analysis of IR spectra (Fig. 6). Similarly to the case of lower oxide, the reaction starts from the first seconds of activation. The shape of the  (W-O) band of WO3 (1000500 cm-1) and the ratio of intensities of their maximums are changed. After activation of the mixture for 1 min, IR spectrum of the sample contains only one intensive band with maximum at 810 cm-1. This band relates to 3 vibrations of WO4 tetrahedrons of the reaction product BaWO4.

X-ray phase analysis showed that the growth of the intensity of BaWO4 phase reflections was accompanied by the decrease of WO3 reflections intensities till their complete disappearance after activation for 5 min (Fig. 7). Microphotographs suggest that the initial stage of the process involves intensive dispersing of particles; their aggregation starts at the second minute of the process. After activation for 5 min, complex oxide particles are formed; the size of their blocks is 2030 nm (Fig. 8).

Fig. 6. IR spectra of initial mixture of barium peroxide with WO3 (a), after its mechanical activation for 10 s (b) and 1 min (c)

Mechanochemcial interaction of BaO2 with MoO3 results in the formation of the complex oxide BaMoO4. The complex oxide with the same composition (BaMoO4) is formed in SHS process between barium peroxide and the higher molybdenum oxide, both reagents being activated preliminarily for 30 s.

Thus, investigations allow us to conclude that the combined potentials of mechanochemical activation and SHS provide extended power for both these methods and enable them to serve as the basis for new resource-saving ecologically safe and waste-free technology of the production of single-phase metastable intermetallic phases and complex oxides.

Fig. 7. X-ray patterns of initial mixture of barium peroxide with WO3 (a), after its mechanical activation for 5 min (b)

Fig. 8. Microphotograph of product of mechanochemical interaction of BaO2 + WO3 (time of MA 5 min)

References

  1. Heinicke, R.Riedel, H.Harenz. Z. Phys. Chem., 227, 62 (1964).
  2. Grigorieva T.F., Barinova A.P., Boldyrev V.V., Ivanov E.Yu. Materials Science Forum, 235-238, 577 (1997).
  3. Zyryanov V.V., in: «Mechanochemical synthesis in inorganic chemistry», ed. by E.G. Avvakumov, Novosibirsk, Nauka, 1991, p. 102-125 (in Russian).
  4. Itin V.I., Naiborodenko Yu.S. High-temperature synthesis of intermetallic compounds. Tomsk State University Publishing House, Tomsk, (1989) (in Russian).
  5. Avvakumov E.G., Potkin A.R., Samarin O.I. A.C. USSR № 975068. BI, 1982, № 43 (in Russian).
  6. Powder Diffraction File. Swarthmore: Joint Committee on Powder Diffraction Standards. Card №№ 8-277, 8-368, 8-372, 5-626.
  7. ibid. Card № 6-0399.
  8. ibid. Card № 17-306
  9. Plyusnina I.I. Infrared spectra of minerals, Moscow, MSU Publishers, 1977 (in Russian).
  10. Lowson L.. Infrared absorption spectra of inorganic substances, Moscow, Mir (1964).
  11. Barker M., Wood D.I. J. Chem. Soc. Dalton Trans., 22, 2448 (1972).
  12. Boldyrev A.I. Infrared spectra of minerals, Moscow, Nedra, 1976 (in Russian).
  13. Landolt-Bornstein H. Physikalisch-Chemische Tabellen, 1951, Teil 2.
  14. Henning O., Paselt G., Z. Phys. Chem., 233, № 3-4, 170 (1933).
  15. Audette R.J., Quail J.W. Inorganic Chemistry, 11, № 8, 1904 (1972).
  16. Grigorieva T.F., Barinova A.P., Ivanov E.Yu., Boldyrev V.V. Doklady Akademii nauk (Reports of the Russian Academy of Sciences). 354, № 4, 489 (1997) (in Russian).
  17. Grigorieva T.F, Barinova A.P, Kryukova G.N., Boldyrev V.V. Proceeding of EUROMAT’97. (5th European Conference on Advanced Materials, Processes and Applications, Maastricht, April, 1997). 2, 333 (1997).
  18. Grigorieva T.F., Barinova A.P., Vorsina I.A., Kryukova G.N., Ivanov E.Yu., Boldyrev V.V. Chemistry for Sustainable Development. 6, 115 (1998).
  19. Grigorieva T.F., Barinova A.P., Vorsina I.A., Kryukova G.N., Boldyrev V.V. Zhurn. Neorg. Khimii. 43, № 10, 1594 (1998).
  20. Smirnov V.I., Aleksandrov V.V. Inzhenerno-fizicheskiy zhurnal. 65, № 5, 594 (1993) (in Russian).

1